CG dinucleotide frequency

[gsgs]

I just made an amazing discovery : the content of CG-binucleotides
in human DNA is much lower (~one fourth) of the content of other
binucleotides, e.g. GC.
This is so basic, fundamental, that it must be known and there should
be a keyword where/how to find more about it ?!
But I never saw it.

The ratio (number of GC)/number of CG) seems to depend on
chromosome-length, it's also seen in other vertibrates,
average ratio maybe 2 ,
not in invertibrates,viruses, bacteria where I got a ratio of ~1.2,
except HIV with 5.8(!) , influenza with 2.2

hepatitis with 1.35 , oryza 1.39 , gallus 5.0 , influenza-B 2.8
human mitochondrial DNA 1.63
human chr01:4.36
human chr22:3.44


so, how is this phenomenon called ? Where to find more about it ?
is it because the assembling of CG takes longer and is thus avoided ?


human chromosome 1:
AA,21387422
AC,11301414
AG,16036982
AT,16766098
CA,16363109
CC,12259077
CG, 2281713
CT,16060830
GA,13448131
GC, 9938922
GG,12248730
GT,11320734
TA,14293281
TC,13465343
TG,16389034
TT,21438891

GC/CG-ratio=4.355904

--------------------------------------------------

OK, I found the answer now myself : "DNA-methylation"

----------------------------------------------------

[Gig77]

Look for CpG islands and methylation. My understanding is that CG dinucleotides are selected against because they are prone to methylation with subsequent deamination.

[gsgs]

thanks. Why is the ratio bigger in longer chromosomes ?

-------------edit--------------
DNA methylation is not found in all species.
It is absent or extremely rare in invertebrates such as
Drosophila and Caenorhabditis elegans, whereas it has
been found in sponges, some insects, sea urchins, and
invertebrate chordates, as well as in all vertebrates and
plants so far examined. (1998)
-----------------------------------